Learning is a highly regulated process in the brain. Such regulation ensures that the specificity for adaptive change is limited to behaviorally relevant contexts, as when movement is modified in response to motor errors. Therefore, neural circuits for learning are typically built with an embedded arrangement of diverse cell-types and connections that support computational flexibility, so that they become responsive for plasticity induction under exceptional circumstances but are otherwise impermeant to change. In the cerebellar cortex, the established framework explaining its function fails to account for how it rejects changes during unexceptional contexts yet maintains the latent ability to induce learning in response to motor errors. In this seminar, I'll highlight my lab's recent work examining the role of synaptic inhibition in regulating cerebellar plasticity induction and learning, discoveries that will help to fill in knowledge gaps regarding how it operates in health and disease.